On December 7 2015 The Royal Society announced that, from January 1 2016, it would require all corresponding authors submitting papers to its journals to provide an Open Researcher and Contributor identifier (ORCID iD). In an open letter published today, seven other publishers – the American Geophysical Union (AGU), eLife, EMBO, Hindawi, the Institute of Electrical and Electronic Engineers (IEEE), PLOS, and Science – joined them, committing to requiring ORCID iDs in their publication process during 2016.
Stuart Taylor, Publishing Director of The Royal Society, kindly agreed to answer some questions about why they – and the other publishing organizations – made this decision, and what he sees as the main opportunities and challenges.
Please tell us a bit about the Royal Society and its publishing program.
The Royal Society is the national academy of science for the UK and Commonwealth. We are an independent charity founded in 1660 by enlightenment scholars to promote science for the benefit of humanity. We started publishing in 1665 when the world’s first science journal, Philosophical Transactions, was born. Today we publish 10 journals across the entire range of science, mathematics, and engineering. We support openness and transparency in science. Authors can publish open access articles in any of our journals have and we have two fully open access journals. Our most recent journal, Royal Society Open Science, operates objective peer review and optional open peer review (which is a popular option for authors and reviewers).
Why did the Royal Society make the decision to require ORCID iDs for submitting authors from January 2016?
We have been supportive of the idea of a universal author identifier for some time and have been displaying ORCID iDs on published journal articles when our authors supply them. But the decision to make ORCID mandatory for corresponding authors arose from the discussions and debates we held in April and May 2015 as part of a four-day event to look at the Future of Scholarly Scientific Communication. ORCID iDs were mentioned frequently at many points over the four days and it became clear to us that there would be great benefits to researchers and to science in general if they were adopted by as many researchers and organizations as possible. We felt that we had a role to play, as a publisher, in driving adoption of ORCID by researchers and to use our influence in the scientific community more generally to spread the word.
What sort of response have you had following your announcement of this requirement in early December?
It’s early days yet, but there was a lot of interest in our announcement. We have had good pickup in various channels and the response so far has been positive. We have updated our author guidelines and linked them to a blog post by our Publisher, Phil Hurst, on the benefits for researchers and we will continue to monitor feedback from our community.
You and seven other publishers are now leading a broader initiative to encourage ORCID adoption in the scholarly publishing process. How did that come about?
Publishers get together often and we are constantly discussing how the system is evolving. I was speaking to a few like-minded publishers at a meeting on open science in Washington and we soon realized that we could achieve more in terms of driving ORCID adoption if we worked together. We agreed on all the key principles; only the timing and precise implementation needed to be figured out, since we all use slightly different systems in our publishing workflow.
What exactly are publishers who sign the open letter committing to?
Publishers are really only signing up to indicate their willingness to introduce the requirement that their authors have an ORCID iD by the time their article is published. The details of which authors and at what point in the publication process they do this is up to them, as is the implementation date. We are really just looking for a commitment and want to put as few obstacles in the way as possible. We are also requiring that the collection of ORCID iDs is done via the ORCID API (authenticated ORCID iDs); that publisher Crossref DOI metadata is updated to include ORCID iDs for authors, so that Crossref Auto-update can be implemented; and that author/co-author ORCID iDs are embedded into article metadata, online, and print versions of the publication. More information about the requirements can be found here.
What does this group hope to achieve through this initiative – both at the organizational and the wider scholarly community level?
ORCID has got off to a flying start, but most scientists are still not signed up. By getting as many publishers as possible on board, we hope to get across a powerful message to the community that ORCID is of great benefit. Those benefits are only fully realized at high levels of adoption and when ORCID support is built in to all the systems researchers use (whether applying for grants, publishing datasets, acting as peer reviewers, applying for appointments etc).
What do you see as the main challenges and how will you overcome them?
Publishers are understandably reluctant to introduce anything that might put off authors from submitting articles and may see the ORCID requirement as just another burden. However, we will be pointing out that the ORCID sign-up takes less than a minute and you only have to do it once! What’s more, many publishers can now update an author’s ORCID record automatically as soon as their article is published, through Crossref’s recently introduced auto-update functionality, which makes life easy for authors. They need only include their iD at manuscript submission (enter once), and grant permission to Crossref to update their record once, and then information about the published work — with their iD — can easily flow into connected systems, such as funder reporting systems and university repositories (re-use many times). We will also be underlining the privacy angle. As so much of our lives is online these days, we all expect to be able to control what information others can see about us. ORCID builds this in so the researcher can select exactly what they display and what they keep private.
Where can Scholarly Kitchen readers find more information about – and sign up for – the open letter?
The open letter is available on the ORCID website, together with the guidelines for publishers about why and how to implement them. We strongly encourage publishers to review them and to consider becoming a signatory. Individuals and non-publishing organizations can also show their support by signing the letter if they wish.
Full disclosure: I am Director of Community Engagement & Support for ORCID. The publishers involved in this initiative asked ORCID to help facilitate communications about their plans, including hosting the open letter explaining their rationale, developing best practices for using iDs in publishing, and maintaining the signatory list.
58 Thoughts on "Why Some Publishers are Requiring ORCID iDs for Authors: An Interview with Stuart Taylor, The Royal Society"
Obviously authors are going to see ORCID as just another burden. But publishers hold most of the cards here – a burden that’s frustrating, but only wastes ~10 minutes of your time per paper is not going to push anyone to another publisher (especially as they’re also imposing unnecessary burdens on you).
Given the alternative is perishing, authors are going to take their lumps to publish.
Brian, I hope authors don’t see this as ‘just another burden.’ There are great benefits for researchers to the ORCID system once it’s very widely adopted. I’d also just like to correct one misapprehension in your post. It doesn’t take ’10 minutes per paper.’ You only have to register for an ORCID iD once (and it takes a lot less than 10 minutes).
“There are great benefits for researchers to the ORCID system once it’s very widely adopted.”
“Once” is the key word there. Publishers have been singing the praises of ORCID for some time, but I’ve been waiting for it to gain traction in the author community, and I’m still waiting. Initiatives like this will drive more authors to sign up, but do authors really understand the benefits they will supposedly receive by doing so? I don’t see anything in the Open Letter that talks about plans to educate authors on how they can use their new ORCID ID, so I suspect many will simply sign up, and then promptly forget about it.
To be clear, I’m not trying to be negative about this – I do see the benefits to ORCID – but author/researcher outreach and education has been lacking since day one in my view, and I don’t see this requirement as doing much to correct that.
Here’s a researcher’s perspective : http://occamstypewriter.org/scurry/2016/01/07/orcids-set-to-bloom-in-2016/
ORCID is not an initiative that authors/researchers asked for, per se. If funders and publishers require it, it won’t matter whether they understand the benefits. They will start seeing the ORCIDs on published papers and be able to see the benefit. Never underestimate the ego factor of authors or their desire to skip steps. If they see that it makes their work easier to find in one place and connect, they will be believers. Historically speaking, the author community hardly ever adopts anything by option. It’s typically by mandate. I am not criticizing them it’s just that being an author is a small part of their day-to-day lives.
And the publishers will include the ORCID iDs in their Crossref deposits and via Crossref’s Auto Update service the author’s ORCID record can be automatically updated with the publication information – and then others can pull that information from ORCID or Crossref.
I can’t speak for other authors, but *I* certainly see it as another pointless burden. It’s something I have to do that takes time. The last paper I submitted, searching around for my ORCID (one of my ORCIDs?), resetting the password, etc., took ~10 minutes (did I also have to find them for my co-authors? – certainly, the amount of time I had to put into overcoming this burden is why I’m very negative about it, rather than just slightly irritated). I expect that’ll be the same next time – with the Royal Society apparently making it mandatory, I won’t really be able to avoid journals that require an ORCID anymore.
And there aren’t any benefits to authors. As far as I can see, the purpose is to offload onto authors organisational stuff that was previously done by publishers? Why would publishers require it? Because it benefits them, but not us.
Disambiguation is an important key to discovery. I have papers published as David Crotty, as David A. Crotty, and as David Alan Crotty. ORCID can help so people looking for my papers without having to know my middle name. The problem is even more exacerbated for people who have common names. Try searching for just the right paper in PubMed by someone named “Smith” or in Google Scholar for a paper by “Wang”. I’m told that in China, the top 100 surnames account for 84.8% of the population. Then there are the researchers who have changed their names, or gotten married/divorced and have papers under different last names or combinations of hyphenated names. ORCID can connect all of these papers written by the same person.
Also, someone can go from the ORCID iD on one of my papers and see more about me, my past positions, my current position, funding and all of the papers I’ve ever written. This can help me get recruited for jobs or help potential collaborators find me. They can go from my papers to realize that I write for a blog and have a Twitter account.
ORCID is rapidly becoming a key tool used by automated systems to keep researchers in compliance with funding agency policies. I’d rather have a system that does this for me than have to do it myself by hand, spending hours/days following arcane rules of my institution and my funder in order to keep my money and my job.
And as I understand it, ORCID will soon become a time-saving tool, allowing applicants to fill in just their number and have other fields of forms (grant submission, paper submission) auto-filled for them (work still apparently in progress).
And seriously, if you can’t keep track of one number, then perhaps you are the person I always end up in line behind at the ATM…
As long as you’re keeping it up to date, sure, people might be able to use your ORCID to see you CV. Which means now you need to keep an up to date CV in five places instead of four.
And it can make me go through steps to ensure I’m doing what my funding agency requires of me. So now this is a task I’m doing in triplicate, rather than just twice. That is not a service to me as an author.
And so on. ORCID doesn’t save any labor for authors, it just takes us from doing n tasks to n+1 tasks.
And, of course I can remember one number. But ORCID isn’t asking me to do that. It’s asking me to keep track of 153 numbers, instead of the 152 numbers I currently have to keep track of. In practice, I can go to the “Dear ORCID, I forgot my password” page, enter three or four of my email addresses until I get a hit, and then google “Brian Lastname ORCID”, and try logging into the two possible accounts and whichever lets me in is me. This isn’t difficult, but it’s certainly irritating.
See Ed’s comment here:
The idea is that by using a small number of broadly accepted standards and persistent identifiers, we can automate many of the tasks you complain about. Use your ORCID iD on your paper, and instead of having to update your CV on ORCID, the metadata of the paper includes that information, which automatically propagates and updates your CV for you.
Similarly, systems like CHORUS (http://www.chorusaccess.org) are built to use these identifiers to automate compliance with funding agency policy. Tag the paper with the ID for the funder and the system is meant to take care of compliance for you, rather than forcing you to do it manually.
And as far as having to remember 153 numbers, the idea here is to reduce that down as much as possible by having all parties agree on a small number of standard numbers. Those agreed-upon and open identifiers can then be used by all parties, again, reducing your efforts (see comment on Single Sign On here http://scholarlykitchen.sspnet.org/2016/01/07/why-some-publishers-are-requiring-orcid-ids-for-authors-an-interview-with-stuart-taylor-the-royal-society/#comment-157319 ).
I understand the idea, and if ORCID worked the way you wished it would, perhaps it’d be helpful. But I’m also aware of the practice, which is that new broad standards aren’t universal, but compete, so you have to do all of them. So it would be great if ORCID reduced that 153, but it won’t. I’ve had to register and submit an ORCID when publishing with one journal, and yet I still had to retain everything else. And my ORCID page, of course, lists nothing more than my name.
If tasks are important, you can’t rely on automation because broad standard covering many different users + automation means occasional failings. So you’ll have to do them by hand *anyhow*. If they’re not important, then ORCID doing them isn’t a favour. So although the idea is great, now I’m making sure I’m compliant with my grant requirements, proving to my university’s enforcement office I’m compliant with my grant requirements (who’re supposed to do that for me in a labour saving like ORCID way, but don’t, and I have to then fix their mistakes), and I’ll have to go through and register it with ORCID again. What I should have to do once, and now have to do twice, I’ll then have to do thrice. It should be obvious why I’m not thrilled.
This job cycle, I joined three broad academic job application processing services, all designed to simplify applying for jobs in bulk (as academics are prone to do). But each app had to be customised, so I went through everything each time anyways, and in most cases I had to send a traditional application as well; the HR departments were saying “You must apply using Vitae”, and the search committee were saying “Send us a normal application”. So, no work was saved, and existing work had to be duplicated. And each service, I think, I used for one or maybe two jobs. So there I had to learn to navigate each service, once, to do a redundant task. When I already knew how to assemble a traditional practice. If Vitae (or whomever) had become a universal standard, used by everyone, then the learning curve could’ve paid back. Just as you’re promising. But of course it didn’t, because that’s not how introducing another standard to be universal works.
All reasonable points, and really the reason why the journals I work with have enabled ORCID but aren’t yet requiring it. It’s early days for all of this, and no, there is no universal standard as of yet (though in my opinion, ORCID stands the best chance of becoming one of all the ID systems I’ve seen). You have to start somewhere.
And I disagree that all the sorts of niggling compliance sorts of requirements must be done by hand. There are now over 700 institutional, funder, and government policies around access to published research papers, all a little different, different embargo periods, deposit in different repositories (or no deposit at all), etc. Nearly every paper has more than one author, most have more than one funding source, and we’re seeing more and more collaboration between researchers at different institutions and more and more international collaboration. As a publisher, I’m not going to carefully go through each and every paper by hand for every author and figure out their individual needs and see that they are fulfilled. I’d like to help, and the only way I can handle the volume of the thousands and thousands of papers we publish is to set up systems that can automate as much of the process as possible.
We want to help lighten the burden for our authors. When we do things like deposit papers into PMC on behalf of NIH funded authors, it is generally very much appreciated. The more we can provide services like this, the better.
See also Cameron Neylon’s recent piece on the benefits of ORCID and why it has been slow to catch on:
“Also, someone can go from the ORCID iD on one of my papers and see more about me, my past positions, my current position, funding and all of the papers I’ve ever written.”
Firstly, this is only true if I log into the ORCID system and enter my employment, education, and funding data manually. A lot of my publications are missing, even using my alternative names.
Second, ORCID hasn’t been widely adopted and isn’t replacing ResearcherID, ScopusID, Google Scholar, ResearchGate, Academia.EDU, institutional repositories, personal website, and/or a written CV. As Brian says, this makes ORCID yet another site containing my CV that I need to constantly monitor and manage. This comic captures it: https://xkcd.com/927/
Hi Emma, I agree that this is a problem, but not really an uncommon one in these sorts of situations (I recently read that there are at least 23 different competing standards for Institutional Identifiers as well). It is also not an insurmountable problem, and the benefits of having open, common persistent identifiers makes it worth solving.
Standards don’t just fall out of the sky. They have to start somewhere, grow in their use and eventually be adopted and required by enough of a community that they do indeed become real standards. It’s easy to complain that we’re not there yet, but given the alternatives (living with 15 plus similar systems, giving up on the promise of persistent identifiers altogether), it strikes me as work worth doing.
I see ORCID as the best contender from the lot you mention above. It’s not tied to a particular company or institution. It is run by the community as a not-for-profit, so the motivation is to serve the community, not to put money in someone’s bank account. It is not part of a scheme to sell advertisements, nor part of a scheme to spy on researchers and sell their data to pharma companies as many of the above clearly are. Those factors to me make it much more viable than any of the choices you’ve listed above.
As has been discussed elsewhere in these comments (with the same link to the same painfully true XKCD comic), it’s still early days, but work is already underway to automate many of the tasks you mention above. Many journal platforms are in the process of implementing a system whereby any paper published with an ORCID iD automatically populates the listing on that person’s ORCID page with the paper, no extra effort required. Similarly, many funding agencies are starting to adopt (and require) the use of ORCID iDs, which will pave the way for automated updating of that information as well.
Shaun, thanks for your comments about the need to better educate researchers about the benefits of ORCID. I completely agree and it’s something we will be focusing on in 2016. We do already have a number of resources available, including downloadable materials on our website (http://members.orcid.org/outreach-resources) and are already working on creating more. Our recent survey gave us lots of valuable feedback about what researchers do – and don’t – understand about ORCID, and which benefits are most important to them, so we’ll be feeding that information into our new materials. Thanks to the Helmsley Trust grant we received in 2015 we are also ramping up our regional workshops, with at least 12 planned during 2016. If you’re interested you can read more about our priorities for 2016 here: https://orcid.org/blog/2016/01/05/orcid-mantra-2016-sustain-lead-and-mature.
It’s great to hear that more end-user outreach and materials are planned for 2016 Alice. My hope is that these materials will not be confined to the ORCID website, but that you will encourage publishers to link to and distribute these materials directly to authors along with the author instructions they already provide, or at the point they are asked to provide or sign up for an ORCID ID (after all, it is in the interest of publishers to get authors to buy into this whole-heartedly). As many have stated, signing up for an ID is quick and easy, but these authors then need to do something with it. If this information isn’t readily available, they will sign up, but they may not stop to think about why this could be important or beneficial for them.
I agree – it’s great that resources are being developed. However these resources not hosted on the ORCID website, they are hosted externally on Slideshare, which is owned by LinkedIn. As far as I can tell (and please correct me if I’m wrong), it’s not possible to download any resources without creating a Slideshare (aka “LinkedIn Slideshare”) account.
FYI, if you already have a LinkedIn account and the email addresses match, Slideshare will log you in automatically and transfer a bunch of personal data from LinkedIn, see https://help.linkedin.com/app/answers/detail/a_id/62999.
Adoption of generally used standards, and ORCIDs in particular, is required for a robust, mature and author-friendly open access future. It took me about 5 minutes to register for my ORCID ID (http://orcid.org/0000-0002-7192-6578). That five minutes would represent a fraction of the time a researcher could ultimately save (in this not yet achieved data driven future) trying to determine e.g., whether her institution has special APC pricing, or whether her chosen journal/license complies with funder requirements. Life will get easier, not harder, as a result of this requirement. Plus, ORCID is a great way to keep you CV up to date.
An immediate benefit of ORCID for authors is single-sign-on across multiple systems. See here for a live demo:
I think the main benefit (is for publishers) to gather contact information of registered users and to spam them with all kinds of ads and announcements for promoting and praising the publishers products.
Then, the contact information could be sold to other buyers and so on for more business of ads and incomes!
For authors, I do not see any ‘real’ benefits.
When an author submits a paper to a journal, they are required to include their name, mailing address, phone number and email address. What makes you think that also including an ORCID iD number would somehow make a massive difference in terms of direct marketing?
“Neither do we ever sell any ORCID data (or give it away!). ”
ORCID might not sell it, but to claim that it doesn’t give it away is highly misleading. ORCID creates an annual Public Data file, that is available for free online under the Creative Commons CC0 1.0 Universal public domain dedication.
See here: http://orcid.org/content/orcid-public-data-file-use-policy
Use of the file is only governed by the “community norms” and essentially pleas from ORCID for people not to misuse the data. Please correct me if I’m wrong, but to “Orchid Farmer (the flower!)” ‘s point, it seems there is absolutely nothing (legally) stopping anyone from extracting names and contact data from the public data file, or from selling that information to another party.
Emma, Thank you for sharing your thoughts!
As to your earlier question about person identifiers. You can connect your identifiers together using your ORCID iD: Scopus Author ID, SSRN iD, ResearcherID, arXiv iD, are among the identifiers you can connect to your ORCID iD. ORCID doesn’t replace these other identifiers, rather it provides an international and discipline-blind identifier key.
Connections between identifiers (for people, places, and your affiliations) enable you to move your research data between systems. For example, when you use your iD during the publication process, the iD becomes part of the paper, just like your name is. Systems that index the paper have enabled search by ORCID iD. If you want, you can use tools to connect your iD to your existing works. Either way, this means other people can search for and find your works more accurately. As more research organizations enable ORCID in their systems, it also means you will spend less time re-entering your research contribution data.
Again, thanks. And please let me know if this explanation is helpful.
“You decide what information in your record is public and what is not”
Sorry, but this does not seem to be true. I just tried to play around the different options available, and when I try to setup the profile to ‘visible to trusted or private’ it always turns back to public!
I am afraid that ORCID is not more than a new way to gather money from naive scientists, in the same way as researchgate, linkedin, and alike… do.
I think that this is a big problem ORCID, or any not-for-profit, community-run organization has to face in a marketplace where commercial interests are flooding researchers’ inboxes with spam in attempts to drive up their stock price or at least their user base to make themselves attractive for purchase. ResearchGate and Academia.edu have really poisoned the well for those looking to work on behalf of the research community.
This is why I am not on researchgate or academia or alike. Already fedup with spams and ads.
Hi ABCDER, you can’t make your ORCID iD private but you should be able to set everything else to private. Our support team (firstname.lastname@example.org) can help figure out any problems you’re having with this so please do contact them. Thanks, Alice
Thanks for your response, Laurel. I’ve spent the best part of a day updating my ORCID iD – linking it to other ids (which of course requires registrations), and entering work and education history. It’s still not a complete record and I can see I’ll need to keep it manually updated (and consistent with my other online profiles and electronic CV) for the foreseeable future.
One issue I’ve found frustrating is in selecting the preferred source for entries with multiple sources. Journal articles display the year, but not critical details such as volume and issue. I cannot easily determine which of those multiple sources is the most accurate or most recent. Some sources have indexed the early online version, some have the final vol/issue. However I cannot distinguish these based on the information on the screen.
A relatively minor issue is why I can only have one entry for “Country”. Is this supposed to be my country of residence, country of current employment or study, or something else? At present these are different for me. In Europe (where I now live), it is not uncommon for people to live in one country but work/study in another (or more). They may desire to work in yet another. Limiting this field to only one is a problem in this respect.
Yes, it seems that different indexes hold somewhat non-overlapping data about the publication. What we are focused on is the persistent identifier (DOI, PubMedID, etc). From our perspective the important thing is to connect your ORCID iD with the publication iD; the specific information about the publication can be resolved from the source document, by the system is connecting with your ORCID record, whether that is a publisher submission system, a university faculty profile system, etc.
Regarding which wizards to use, we generally suggest that a goal is to attach your ORCID iD onto your existing publications in the systems that people in your field/region/etc. use to search and discover. That makes all of your works more easily and accurately found. There are several wizards available through the ORCID interface or in other abstract/index systems that support this “attach iD to existing papers – push data to ORCID” process. For your new papers, use your iD when you submit (or at acceptance if you are a co-author) and then approve Crossref as a trusted party so your paper information (including the DOI) can be posted to your record w/o further manual intervention.
As mentioned earlier in the comments, one of our priority areas this year is to provide more information for researchers on how to use ORCID. Our current docs are available here: http://support.orcid.org/knowledgebase/topics/32827-website-user
I hope this response helps in the meantime. And please do contact our support desk (https://members.orcid.org/contact) if you have a specific issue.
I read with interest this post, as well as the last one here from Alice Meadows: http://scholarlykitchen.sspnet.org/2015/12/15/all-about-orcid/. As the editor of a large journal in a major surgical subspecialty, the identity of authors who send their work to us matters deeply to me. I enjoy the Scholarly Kitchen blog, but am finding the presentation about ORCID to be somewhat one-sided.
Although ORCID offers many potential benefits, it does not achieve the purpose for which it was named; the last-two letters of its acronym and proponents’ protestations to the contrary notwithstanding, ORCID does not reliably identify that an author is who (s)he claims to be.
In Europe and the United States, the idea that someone would share an ORCID identifier (iD) seems incongruous. In China, which is producing the largest year-over-year increases in manuscript submissions at many journals, identity “sharing” seems commonplace. We have found that it is routine practice for multiple authors to share single email addresses, and for individual authors typically retain (and submit work from) multiple email addresses. The latter is no malfeasance; it is a practical solution to spotty email service in many parts of mainland China. However, the former represents a substantial problem for journals who wish to have authors – in particular corresponding authors – uphold the responsibilities of authorship as articulated by the ICMJE, COPE, and others.
This problem is exacerbated when multiple authors also share an ORCID iD, which we also have identified, and recently reported to ORCID. ORCID has confirmed this instance of subversion of its raison d’être by these authors. Interestingly, we found no formal process to address identity fraud on ORCID’s website (we reached out to ORCID through its “contact us” link), and – much more importantly – no mechanism to prevent this kind of fraud from occurring when individuals sign up for the service. Any system that is based on an individual asserting his or her identity by email alone is vulnerable, both to fraud and to hacking. We have observed both fraud and hacking in submissions using email identifiers, including, as noted, misrepresented identity using an ORCID identifier.
It is anathema to most of us writing from Europe and North America that someone would share an ORCID account with another individual, so that when one of them writes, the other gets the credit. By contrast, this is NOT an unusual concept by any means in China, where our interviews with authors and editors have found that younger faculty members routinely write on behalf of their more-senior colleagues, provide ghost authorship for the elder “author” – in the instance we described, facilitated by ORCID – and gain professional advancement through the transaction.
Two years ago, I would have considered a story like this some kind of conspiracy theory. After the reviewer-fraud debacle that has resulted in dozens or hundreds of retracted articles (see: http://www.nejm.org/doi/full/10.1056/NEJMp1512330?af=R&), after our own inquiries to Chinese authors and editors, and some electronic sleuthing that demonstrated the alarming frequency with which authors share email addresses (and now, seemingly, an ORCID iD), I’ve reluctantly concluded that the phenomenon is real, and journals – including ours – must attend to it. By no means am I suggesting that ORCID has caused this problem, but neither can editors or grant-funding agencies assume that ORCID will stop it.
Really interesting point Seth. How much of a responsibility does the group behind any community standard have toward policing the use of that standard? I think there’s often a laissez faire approach, create the tool, put it out there and let people do what they will do with it. And as you note, that can lead to abuse.
But is the sort of identification and fraud prevention that you’re seeking really the mission of a group like ORCID? Their tool is more about disambiguation and for use in automated systems where it would be hard to see much to gain by fraud. For the peer review fraud you mention, I’m not sure ORCID is really meant to be a policing mechanism for this sort of identification. Whether they supply an ORCID iD or not, you’re still going to have to verify that they are who they say they are (and that their email address is correct).
Should we allow anyone who wants one to get a DOI for something they put online or should they have to apply and be carefully vetted? Should FundRef only accept funding agencies that meet certain criteria or to try to be as inclusive as possible?
I don’t have an easy answer for you, other than that a carefully regulated, moderated system is vastly more expensive than an open system.
David, thanks. I don’t have easy answers, either. We are deploying considerable resource in our office to run down this problem with each manuscript we receive, and the more we look, the more common the problems appear to be.
But to the degree that one purpose of ORCID is disambiguation, we are concerned (and again, have observed) that this mission is not reliably achieved.
You are surely correct that fraud prevention may be too much for ORCID to take on. An issue, though, is that an ORCID iD on a paper – again, look at the initials in the acronym – gives the typical reader (who may know little about ORCID) the impression that identity has been verified, when in fact nothing of the sort has happened.
This is not a problem for the many people who are using ORCID to support their own grant applications and the like, and ORCID has many beneficial features. But all journals are wrestling with how to manage the concerns that have arisen from China (and somehow the latest directive from Beijing seems less than reassuring: http://bigstory.ap.org/article/485c9325497e4114abc530d41308e28c/china-issues-rules-banning-dishonesty-science-publishing ), and given that China is among the fastest-growing countries in terms of numbers of manuscript submissions, these concerns are pressing.
Adding requirements on authors – such as a requirement to use ORCID to submit a manuscript – adds burden to a system that many authors already find burdensome. ORCID appears to facilitate (or at least lend an air of validity to) a pattern of misbehavior that we already are having enough trouble managing.
t this moment, I am not tempted to make it a requirement of our authors.
This does highlight what ORCID is and what ORCID is not. ORCID is a system for providing unique identifiers to individuals so as to disambiguate both individuals and their contributions. ORCID does not certify or validate identity, thus opening it up to fraud, redundancies, hacking, and human error. The problem is not that ORCID does not validate identification, the problem is that it is not obvious that ORCID does not validate identification.
While the technical architecture of DOIs and ORCIDs are similar, I don’t think a DOI is a good comparison for ORCID as with a DOI there is no inherent claim to authenticity. A DOI links to a resource. The resource is what the resource is — there is no claim that the resource has somehow been validated (that is what CrossMark is for – a validation layer that sits on top of the DOI infrastructure).
Because ORCID is about unique identification, and there is not a “validated individual” icon, like with Twitter accounts (or a separate validation system, like with CrossMark), users might assume that some form of validation has happened.
And of course validation could be baked into ORCID registration if ORCID chose to do so. ORCID could require some for of “two-factor” validation. They could require two independent checks on a user’s ORCID registration, issuing only a provisional ORCID until the second validation. For example, one might be publication in a journal where the journal can vouch for identity (as the author may be known to the editors). A second validation could be publication in a conference proceeding or reviewing for a journal where or some other professional activity where the user is known to the validating party. This would all be possible but would significantly constrain adoption of ORCIDs, perhaps to an extent that it would not be able to reach the critical mass necessary to become a common standard.
There is a balancing point here between ease of adoption and more robust validation. I will not pretend to know where that balancing point is, but it should be explicitly acknowledged that ORCID is focused on adoption not identity validation at this point and that an additional layer of validation maybe a needed piece of industry infrastructure.
Michael, my compliments. You win the “line of the day” award: “The problem is not that ORCID does not validate identification, the problem is that it is not obvious that ORCID does not validate identification.”
You nailed it. And it is not ORCID’s job to take on any project, responsibility, or task that its board does not wish to take on.
But as you point out, any relationship between an ORCID “iD” and an individual’s “identity” is unverified. As you also point out, in the absence of certification or validation, the process is susceptible to “fraud, redundancies, hacking, and human error.”
All that being so, I am confused by your second sentence – which for me is the crux of the matter – how does a product that provides unverified identity disambiguate anything?
Seth – I meant in my second sentence that ORCID disambiguates identity for the 99% of people (or whatever the percentage actually is – maybe it is much lower) that are using the system honestly and as it was intended. The system works to tie together my works published under Michael Clarke, Michael T. Clarke, MT Clarke, and so forth and to clarify which ones are mine and which are authored by other Michael Clarkes. It helps with name changes (marriage, divorce) and so on. It is not foolproof (and there are many fools) but seems an improvement over not having any unique identifier at all. That said, it will ultimately depend on how big a problem misuse of ORCIDs become. If the number of people sharing ORCIDs, forgetting their ORCIDs and just registering for new ones, or fraudulently using ORCIDs rises above a certain level, the utility of the system will, of course, be compromised.
And again, Michael, you have hit the spot. The benefits you cite are the reasons that as an individual I possess an ORCID iD and use it (when I can remember to). Its important shortcomings are why, as an editor of a large journal, I don’t require it of our authors.
Since this Scholarly Kitchen article dealt with ORCID becoming a requirement for publication across many, many journals — without any mention of its shortcomings, despite it being seemingly misunderstood as a validator of identity and easily subverted by those so inclined — it felt important to speak up.
I have never replied to a blog post before…seem to have leaped in with both feet on this one. I hope this does not become a habit. I have a day job…
Malicious or malevolent persons can harm you with or without ORCID even in the real life, they can usurp your name, your passport, your car registration documents, etc…, so the point you are raising about ORCID and validation does not make much sense because it is applicable to everything in real and virtual life (internet).
The problem in my view is not about the validation of author but in the usefulness of the ORCID. If ORCID proves to be useful or beneficial for authors with time, all the rest is a matter of common sense and trust like in any other online service or real life.
When you created your email, which validation was applied? Nothing, and, even though emails are now commonly used for more sensitive process (eg. online payment) than for submission of manuscripts.
Also, don’t forget that all accounts, websites, etc are subject to fraud and scams are not proper to ORCID only.
If ORCID should be adopted and developed, it should remain simple and devoid of any advertising activities. That’s what would make it ‘useful’ and attractive for authors but not the kind of burdensome and compelling things you are talking about.
Why an author would ‘steal’ the ORCID of another to submit a manuscript?
Even if this would happen, after submission confirmation emails are sent to the ORCID linked account, so any fraud should be unveiled.
One would ask, then, how an individual person’s identity might be verified to the extent that such a publisher would be satisfied?
To use the China example, a person is supposed to use their personal identity card to, e.g., register for certain internet services, or to get access in an internet cafe. Yet even here identity information is shared, for whatever reason. Must the person then submit a manuscript in person, complete with two forms of photographic identification?
You ask a good question, On Sanma, although I think your note is a bit facetious, at least at the end. Even so, it is a good question.
I’ll turn it back on you – what do you think a journal (or its readers) should reasonably expect from authors in terms of the authenticity of their identities?
The International Committee of Medical Journal Editors (ICMJE) – whose criteria are as thoughtful as any of which I’m aware – says that authors are responsible for (at least) four things:
“-Substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work; AND
-Drafting the work or revising it critically for important intellectual content; AND
-Final approval of the version to be published; AND
-Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.” (See: http://www.icmje.org/recommendations/browse/roles-and-responsibilities/defining-the-role-of-authors-and-contributors.html)
It seems to me that none of these conditions can be met by an “author” who misrepresents his/her identity or shares an ORCID iD when (s)he submits her work for publication. Ghost authorship, guest authorship, and other kinds of identity fraud all likewise violate ICMJE’s terms.
The ICMJE further says that the corresponding author is “…the one individual who takes primary responsibility for communication with the journal during the manuscript submission, peer review, and publication process, and typically ensures that all the journal’s administrative requirements, such as providing details of authorship, ethics committee approval…are properly completed…The corresponding author should be available throughout the submission and peer review process to respond to editorial queries in a timely way, and should be available after publication to respond to critiques of the work and cooperate with any requests from the journal for data or additional information should questions about the paper arise after publication…” (Same webpage as above).
I would submit that none of those conditions can be met if the journal can’t tell whether an individual is whom (s)he says (s)he is because of shared email addresses, fraudulent submission processes through third parties (see any of several news stories on this from last year), or ORCID iD’s that are used by more than one author.
Don’t you think the journal should know with whom it is dealing during the review process, in case questions arise? Don’t you believe the reader is entitled to know who is responsible for the work published in the pages of a trusted source?
To be clear: I am not at all suggesting that ORCID caused this problem. It did not. I use ORCID myself as an author, and consider it a nice convenience for individuals going up for promotions or seeking grants. My point is simply that to the casual reader, ORCID appears to verify or disambiguate identity, when in fact it does not necessarily do so. For that reason, I do not now require that authors use ORCID when they submit their work to the journal that I edit.
Seth, More and more, universities and other employers of researchers are using ORCID to assert affiliation. This will make it possible for authors to provide that assertion to those journals who are requesting their ORCID iD. Some of the side benefits to this: it streamlines the naming of organizations by using an organization identifier; and (if journals start to use these organization identifiers) that in turn will improve the accuracy of organization name-based publication searches. So, it is not just about using ORCID iDs for researchers, but more generally using persistent identifiers as a means to support the unambiguous exchange of names in research information systems.
Is there any information on the ORCID site explaining how organizational identifiers work to verify affiliation? It’s not clear to me how “current” organizational / institutional information is linked to an author’s record, such that a journal could use it to verify affiliation. Is information provided by an organization or institution distinguishable from that entered by author?
More importantly, what happens to the information when an author leaves the organization or institution? Are historical, verified affiliations maintained? In terms of publishing, it’s not unusual for authors to be submitting manuscripts and/or navigating them through the publication process, despite an affiliation ending or changing. If ORCID can’t handle these realities of academic employment, this is a big, big problem.
Emma, re: organizational affiliation. Each name in the list is backed up by an identifier. You can self-assert (i.e., pick a name from the list and add to your record) as I see you have done. Our doc on the affiliation user interface is here: http://support.orcid.org/knowledgebase/articles/287018-add-education-employment-website-user
Your University is an ORCID member, and you can also (soon) attach your ORCID iD to university systems. There will be a communication from the university about this coming in the next month or so, I expect. That will allow the university, with your permission, to post their assertion of your affiliation and also to update that when you depart. I suggest you contact your library for more information about timing. Here is a link to your library’s guide on ORCID: http://guides.library.uq.edu.au/research_impact/author-identifiers.
Quote from Alice Meadows:
“The only information required to set up an ORCID iD is one version of your name, a valid email address, and a password – and the default setting for email address(es)
is always private, so that information won’t be shared without the record-holder’s permission”.
This sounds OK, but the problem is in the “forcing character” by publishers to
have an ORCID to submit papers. This is an anti-democratic and anti-scientific approach.
Let authors free to try it and then see if could be beneficial for them.
With time and testing, authors would eventually see if ORCID could be beneficial, and if so, they will
adopt it progressively. Forcing people is never a good practice.
Alice, how would this part
…We are also requiring that the collection of ORCID iDs is done via the ORCID API (authenticated ORCID iDs)….
work? If I understand correctly, ORCID for each person has to be confirmed via log in. This would mean that each co-author (including those Profs who hardly know how to switch on a computer:)) would need to login to ORCID and confirm it before publication. It will most likely delay the publication…
Hi Aliaksandr. Publishers signing the open letter are committing to requiring an ORCID iD for only the corresponding author. While there would be a substantial benefit to authors and the research community if co-authors were to connect their ORCID iD to submitted papers, there is still work to be done to streamline the process. Publishers and the vendors who support the manuscript submission process are working to optimize collection of ORCID iDs from co-authors.
Thank you, Alice. This clarifies the mechanism and leads me to think that one would need some infrastructure to “claim” one’s publications co-authored as a non-corresponding author (and tell that to ORCID/Scopus/ISI/Publishers about this). Dead authors or those not using Internet would be a problem in this scenario…
I think it is interesting — and important — that Stuart mentioned that The Royal Society requires ORCIDs prior to publication — i.e., not on submission. This seems to me like a good way to be sure there are not barriers to submission.
Perhaps The Scholarly Kitchen would consider requiring ORCID for comments? (Or is the possibility of anonymity essential?) This comes under the concept of “eating one’s own dog food”.
On a practical note, does WordPress support the use of ORCID iDs? On a philosophical note, as I’ve said elsewhere in these comments, I lean more toward encouraging use, rather than requiring it. And as you note, there’s a difference between publishing one’s own work and leaving a comment on the work of others. Just as I feel there’s a need for anonymity in peer review in order to allow one to speak truth to power, there’s a similar value here.